They are not new anymore, yet you don’t see them absolutely everywhere. Search clouds are visual representations of your website’s most popular searched keywords. Search clouds first appeared on the popular photo sharing Web 2.0 website Flickr.com in the form of tag clouds (which I shall discuss later). I like them for a lot of reasons. As a webmaster, I like them because they tell me what my visitors are doing on my website. I can leverage useful information in a snapshot from a search cloud to judge what the interest areas of my website are. By acquiring this information, I can exploit the incoming traffic in so many ways. I can take advantage of popular search queries by:
- plugging in relevant advertising efforts into most popular content areas.
- creating new content for tags that are popular to get more authority on a subject.
- diverting traffic to other interest areas of the website.
- optimizing content from user statistical data through a user experience learning curve.
- finding and evaluating my niche audience.
- keeping content up to date.
- producing spider bait for search engine bots.
You may argue that I may get those statistics via an analytics tool, I agree, but analytics tools show you most popular pages, events, trends, visitor history and search queries of external sources. What you get with a search cloud is an overview of most popular “internal” searches conducted on your site according to word frequency. You could however use your code wizardry to build reports from saved search queries.
Apart from the data that a search cloud provides you, you should also consider what value it offers to your audience. Here are a few advantages:
- a bird’s eye view of what’s hot and happening on your site.
- a fast and readily available method of searching through your content.
- a first point of reference to your loyal visitors.
- maximizing user experience and performance of your site.
There is another flavour of the cloud offering and its called the tag cloud. The difference is simple. Search clouds comprise of search queries weighted by the size of a font that are generated dynamically through user activity on your site. Tag clouds, on the other hand are the most referenced words in your metadata within your content that you define yourself. I prefer search clouds as you are not enforcing your opinions on your audience rather putting the audience’s needs and wants first and foremost and creating a self generated user centric tool on your site. A shortfall of search clouds is that you need data acquired via user activity before you can build one. If your website is relatively new, you may find that tag clouds work better for you as you can probably ascertain by looking at my own tag cloud on the right hand navigation pane.
Examples of search and tag clouds
One of the most famous clouds is from the Web 2.0 school of thought.
A US based company came up with a novel idea of creating a cloud search engine. You define your queries and give them importance by weighting them and then running a search. Give it a go at SearchCloud.net
The everyday phenonmenon at del.icio.us
How to develop a search cloud
Clouds come in different shapes, sizes and frequencies. They work differently for each type of site or industry. The functional workings of a cloud are usually the brainchild of the webmaster or technical architect of the website. You can define what your search cloud does for you and your business. I would recommend some research and development to come up with a solid technical specification. Some best practices on design and aesthetics of search clouds should also be considered before you develop a custom-built search cloud. Help is available all over the web. There are numerous open source applications that will help you create a search cloud for your website for free too.
Customize your CMS application to build your cloud
An idea I came up with when I was the technical architect for a client’s website was to build a few scripts around the search cloud. Configuring your website/CMS to store all data of your visitors keyword search activities is the first obvious step. Separately, you can build a keyword bank in your database to act as an automatic moderation tool for your search cloud to flush out non-relevant keywords. In practice the script would screen any words against your keyword bank to find matches. Once a match is found, the keyword is allowed to be automatically published in your cloud. The frequency of exact matches would determine the size / weight of the font.
Use analytics tool to build your cloud
Why not go a step further while you are at it. Why not combine most popular keywords coming from external search engines to your search clouds to give them more weighting. This may easily work for you or against you, so be careful. I will let you be the judge of that. Almost all analytics tools give you reports on what keywords your users used to find your website. This data can be collected from either your analytics application’s front end, an exported report or even its database. You can configure your website to talk to your analytics tool’s database to pull that data in at timed intervals for auto publish activities. You may however not have database access to any third party hosted analytics tools like Google Analytics, but you may get your hands on this juicy data from licensed applications such as WebTrends, Omniture and Speedtrap.
Search clouds are here to stay
More and more people are warming up to search clouds nowadays. It is becoming a buzzword and the first thing people associate to Web 2.0. Perhaps you will see search clouds making their way into usability guidelines and best practices for websites.
My research has concluded that you will find hundreds of websites talking about tag clouds, but only a handful explaining search clouds. Just my two pennies worth.
Please feel free to leave a comment with links to your search clouds.1